New York, January 17: Artificial Intelligence has huge potential to resolve the world's pressing issues and come up with innovative solutions. However, the the AI GPT models are learning about information related to different aspects of the world such as society, businesses and human nature. The information generated by the AI models is unique and deemed free from copyright, the models are trained based on the copyrighted material of other companies.
According to the report by The Verge, a non-profit group called Fairly Trained plans to "certify AI models that ask permission to use the copyrighted material.". The report mentioned Ed Newton-Rex, former Stability AI vice president for audio, founded the group. As per the report, the Fairly Trained - "adds a label to companies that prove they asked for permission to use copyrighted training data. Microsoft and OpenAI Sued by The New York Times for Copyright Infringement, Lawsuit Calls Companies To Destroy AI Chatbot Models Using Copyrighted Material: Report.
The report mentioned that the organisation will award companies that license protected data to train its models. This first accreditation by Fairly Trained will be called the 'License Model' certification. However, the report said the company would not issue this certification to the developers relying on the fair use argument to train models.
Fairly Trained mentioned in its official blog that there is a divide between the two types of Gen AI companies. It further said that two types include the companies that get consent from training data providers and those that do not. The blog further mentioned that Fairly Trained has already certified nine generative AI companies involved in music, singing, voice generation, and image. These certified companies are - BRIA AI, Boomy, Beatoven.AI, Soundful, LifesScore, Tuney, Somms.ai, Rightsify, and Endel. Meta CEO Mark Zuckerberg To Participate in Deposition as Part of Ongoing ‘Facial Recognition’ Technology Lawsuit in Texas.
As per the The Verge report, it is a "fraught" issue in generative AI to train AI models with copyrighted data. Regarding this, there was also a lawsuit filed against OpenAI and Microsoft by The New York Times, saying the companies used copyrighted articles to train their AI GPT Models without permission and payment. The lawsuit also reportedly said to destroy AI chatbot models using copyrighted material.
(The above story first appeared on LatestLY on Jan 17, 2024 11:05 PM IST. For more news and updates on politics, world, sports, entertainment and lifestyle, log on to our website latestly.com).